As a member of the herpesvirus family, cytomegalovirus (CMV) establishes a life-long latent infection in the human host. The virus can be reactivated to produce serious disease in immunosuppressed patients and in recipients of blood and organ transplants. Progress in defining which cells are latently infected with CMV and in elucidating the mechanisms involved at the molecular level has been slow, although recent advances have made through the application of recombinant DNA and monoclonal antibody technology. Relevant and extremely important information concerning this issue has been derived over the past 10 years in a well-defined murine CMV (MCMV) model, which closely mimics the features of latent human CMV infection. In this proposal, we will apply molecular probes in the study of the latent infection of spleen cells from which the virus can be reactivated in vitro. Specifically, the spleen cell subsets which harbor latent MCMV will be identified as definitively as possible whether or not the virus can be induced to replicate. Secondly, we will define the extent of viral gene expression in terms of RNA transcription in relevant spleen cells which are latently infected. Third, we will characterize the events at a molecular level which occur prior to and during reactivation of latent MCMV in vitro. Similarly, these events will also be carefully dissected as the latent infection is established initially in spleen cells of mice in vivo and in vitro. Finally, other cells and tissues relevant to the issue of human organ transplantation will be probed to determine which specific cells are latently infected using in situ DNA hybridization. These experiments will delineate fundamental aspects of latent CMV infection which cannot be fully characterized in humans.